ADB Economics Working Paper Series

Transcription

1 ADB Economics Working Paper Series Asian Capital Market Integration: Theory and Evidence Cyn-Young Park No. 351 June 2013

2 ADB Economics Working Paper Series Asian Capital Market Integration: Theory and Evidence Cyn-Young Park No. 351 June 2013 Cyn-Young Park is Assistant Chief Economist at the Asian Development Bank. An earlier version of this paper was prepared for the conference entitled Asian Capital Market Development and Integration: Challenges and Opportunities, which was convened on 24 October 2012 in Seoul, Republic of Korea and sponsored by Asian Development Bank, Korea Capital Market Institute, and Peterson Institute for International Economics. The author would like to thank Rogelio V. Mercado Jr. for his excellent research assistance, as well as conference participants for their comments and suggestions, as these greatly improved the paper.

3 Asian Development Bank 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines by Asian Development Bank June 2013 ISSN Publication Stock No. WPS The views expressed in this paper are those of the author and do not necessarily reflect the views and policies of the Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. By making any designation of or reference to a particular territory or geographic area, or by using the term country in this document, ADB does not intend to make any judgments as to the legal or other status of any territory or area. Note: In this publication, $ refers to US dollars. The ADB Economics Working Paper Series is a forum for stimulating discussion and eliciting feedback on ongoing and recently completed research and policy studies undertaken by the Asian Development Bank (ADB) staff, consultants, or resource persons. The series deals with key economic and development problems, particularly those facing the Asia and Pacific region; as well as conceptual, analytical, or methodological issues relating to project/program economic analysis, and statistical data and measurement. The series aims to enhance the knowledge on Asia s development and policy challenges; strengthen analytical rigor and quality of ADB s country partnership strategies, and its subregional and country operations; and improve the quality and availability of statistical data and development indicators for monitoring development effectiveness. The ADB Economics Working Paper Series is a quick-disseminating, informal publication whose titles could subsequently be revised for publication as articles in professional journals or chapters in books. The series is maintained by the Economics and Research Department. Printed on recycled paper

4 CONTENTS ABSTRACT v I. INTRODUCTION 1 II. DE FACTO VS. DE JURE FINANCIAL INTEGRATION 3 III. CAPITAL MARKET INTEGRATION IN EMERGING ASIA 7 A. Integration of Stock Markets in Emerging Asia 7 B. Integration of Bond Markets 11 C. Degree of Financial Integration and Extent of Spillover of the Negative Impacts of Regional and Global Financial Crises into Emerging Asia Economies 13 IV. CONCLUSIONS AND RECOMMENDATIONS 23 REFERENCES 27

5

6 ABSTRACT Financial integration is a multidimensional process through which allocation of financial assets becomes increasingly borderless. This paper assesses the progress achieved thus far in capital market integration in Asia, and compares regional capital market integration with global financial integration. The results of the analysis on which the paper is based indicate that while the pace of regional integration of financial markets in Asia's emerging economies has accelerated in recent years, these markets remain more integrated with global financial markets than with other financial markets in the region. Further, integration of the region's domestic local-currency bond markets with their regional and global counterparts lags the pace of integration of its equity markets. The study also assesses the degree to which volatility in equity- and bond-market returns driven by financial turmoil originating at both the regional and global levels spills over into emerging Asia domestic equity and bond markets. The results of this analysis indicate that such spill-over significantly impacts both domestic equity and bond markets in the region. This finding suggests that ongoing regional capital market integration initiatives should take into account the risk of contagion that regional financial integration presents, and introduce measures for mitigating such risk as a means of ensuring financial stability in the region. Keywords: capital markets, emerging Asia, regional financial integration JEL classification: F30, F36, G15

7

8 I. INTRODUCTION As articulated by Cavoli, Rajan, and Siregar (2004) in their survey of East Asian financial integration, financial integration is a multidimensional process closely associated with development of financial markets. It is thus unsurprising that the deregulation of the financial sector in Asia that began during the 1980s and resulted in a significant increase in cross-border financial transactions ultimately led to the region's ongoing financial integration. During the period following the Asian financial crisis of 1997, many Asian economies modernized their financial sectors and strengthened linkages with the financial sectors of other economies in the region. This has led to considerable maturation of many of the region's domestic capital markets, its local-currency bond markets in particular. The soundness of the region's domestic banking systems has likewise improved, in that domestic capacity for financial supervision has become more sophisticated. Intra-regional financial sector policy coordination has likewise strengthened, as demonstrated by the ongoing development of regional arrangements for macroeconomic monitoring and liquidity support. Nevertheless, there remains significant variation in the degree to which the region's domestic capital markets have matured and integrated with others in the region and beyond. Economic theory suggests that financial integration brings with it significant benefits, including lower costs of trading financial assets, more diverse investor portfolios, and more stable consumption patterns, particularly during periods when the level of economic activity fluctuates widely. Given the absence of restrictions on capital mobility, financial integration allows the level of domestic investment to no longer be constrained by the size of the domestic savings pool, since integration allows foreign capital to be used to underwrite domestic investment. This appears to be an important feature of financial integration, given the direct, positive relationship between domestic savings and investment confirmed by publication of Feldstein and Horioka's seminal paper in 1980, as well as similar literature that followed in its wake. While more recent studies report a weakening in this relationship in advanced economies due to international financial integration, the Feldstein Horioka puzzle of why investment remains linked to domestic savings remains unresolved, the conventional wisdom of financial globalization that has developed over the past few decades notwithstanding. Economic theory suggests that absence of restrictions on capital mobility increases the degree of efficiency with which international financial resources are allocated. The rationale for this view is that capital should automatically flow from capital-abundant to capital-scarce countries, the returns to capital being higher in the latter set of countries than in the former. In turn, the rationale holds that capital flows of this nature supplement the domestic savings pool in capital-scarce countries, thereby allowing domestic investment in such countries to increase and economic growth to accelerate. However, the empirical evidence concerning this matter suggests quite the opposite. It in fact confirms that Asia's considerable net savings tend to flow to capital-abundant countries rather than to capital-scarce ones as the theory would predict. In the wake of Lucas' now-famous 1990 paper, this disparity between what economic theory would predict and what actually occurs is often referred to as the Lucas Paradox or the Lucas Puzzle. Many economists including Lucas himself have offered explanations as to why capital fails to flow from capital-abundant to capital-scarce countries. These explanations fall into two broad groups (Alfaro, Kalemli-Ozcan, and Volosovych 2005), the first of which focuses on differences in fundamental economic variables. Examples of the latter include the degree of technological development, presence or absence of particular critical factors of production, or

9 2 І ADB Economics Working Paper Series No. 351 differences in the level of the output capital price ratio, government policies, or institutional quality. Ultimately, this set of explanations attempts to provide a rationale for the real-world fact that some countries attract more foreign investment than others. The second group of explanations focuses on capital market imperfections, and the differing stage of development of financial markets in advanced and emerging economies (Martin and Taddei 2012, and Matsuyama, 2007). The theoretical foundation for this group of explanations is that capital flows respond to the risk-adjusted rate of return to capital rather than to its corresponding nominal rate. Thus, the higher level of risk associated with the nominal expected returns to capital in capital-scarce countries prevents real-world flows from corresponding to those predicted by the theory. In other words, once adjusted for risk, the difference in the rate of return to capital in capital-scarce countries as compared to capitalabundant countries is simply insufficient to cause capital to flow in the direction suggested by economic theory. Further, because of the sheer volume and quality of information transmitted by credit markets in advanced countries, more capital flows into such economies than would otherwise be the case. To sum up thus far, while the volume of cross-border capital flows has increased substantially since the mid-1990s (Figure 1), the paradox that articulates the inconsistency between the direction of capital flows predicted by the theory and what actually occurs remains largely unresolved. Indeed, Gourinchas and Jeanne (2007) find a natural tendency for capital to flow toward countries that invest and grow less than other countries. This phenomenon was dubbed the allocation puzzle by Prasad et al. (2007), as it flies in the face of the neoclassical view that differences in productivity growth determine the international allocation of capital. Prasad, Rajan, and Subramanian (2007) also find that developing countries that invest more and rely less on foreign capital grow faster than countries that invest more, but rely more on foreign capital as a funding source for financing such investment. They suggest that even in faster-growing developing countries, the capacity for absorbing foreign capital may still be limited, either because of underdeveloped financial markets, or because of a natural tendency toward overvaluation of domestic currency resulting from rapid capital inflows. Ultimately, the recurring episodes of financial turmoil that have occurred over the past 2 decades illustrate that increasingly integrated financial markets have a tendency to experience financial crises, and that the negative impacts of such crises tend to spill-over into the markets in which financial assets are traded. In this regard, there is little doubt that abrupt swings in global investor sentiment have impacted the performance of Asia's domestic equity markets over previous decades. For example, during the global financial crisis of , emerging Asia s equity markets experienced sharp reversals in foreign portfolio investment inflows as a result of deleveraging of global financial institutions, as well as a sharp increase in investor risk aversion, this being particularly true of markets in which foreign investor participation had been relatively high. Such outcomes suggest that the benefits of financial integration predicted by economic theory have somehow not been fully achieved. There are a number of reasons why this may be so. For example, financial integration at the global level is incomplete. This may be one reason why the actual pattern of international capital flows diverges considerably from that predicted by theory as noted above. Further, financial integration often brings with it increased risk in the form of volatility of asset prices and then financial market returns, as well as abrupt reversals in capital flows. Such outcomes which are familiar to many emerging market economies underscore the importance of appropriate financial supervision when moving the domestic

10 Asian Capital Market Integration І 3 economy toward greater overall financial openness, and loosening of restrictions on capital accounts in particular. In this regard, it is important for policymakers to understand that there are varying degrees of financial integration, each of which has implications for stability of the domestic financial system, as well as the ability of the domestic economy to absorb shocks such as increased volatility in financial market returns or reversals in international capital flows. Finally, particularly in cases in which financial integration is pursued in a deliberate and methodical manner such as in emerging Asia the progress achieved tends to be far from uniform across individual economies, as well as across the various subsectors that comprise the overall domestic financial sector. This study has two major objectives. First, it reviews the progress emerging Asia has achieved thus far in capital market integration. Second, it assesses the degree to which the financial integration that has been achieved has increased the vulnerability of the region's individual economies to the negative impacts of financial crises generated elsewhere in the region or at the global level. The overall goal of such analysis is to better understand how the costs to the domestic economy associated with financial integration may be minimized, so as to maximize the net benefits potentially available from financial integration. From a policy-making perspective, a key issue confronting regional financial cooperation and integration is how best to shape national and regional policies in a way that allows the region's individual economies to maximize the potential gains from financial integration, while at the same time minimizing exposure to the risks associated with it. This paper is organized as follows. Section II assesses the degree to which de facto as opposed to de jure financial integration has been achieved in emerging Asia. In this regard, Section II uses several widely-accepted quantitative indicators to highlight recent developments in financial integration at the regional and subregional levels. Section III then quantitatively estimates the degree to which actual integration has been achieved in equity and bond markets in emerging Asia. Section IV uses principal component analysis to assess the extent to which financial crises generated at the regional and global levels have impacted the financial markets of individual economies in emerging Asia. The conclusions and policy recommendations presented in Section V then complete the paper. II. DE FACTO VS. DE JURE FINANCIAL Integration While there is no universally accepted definition or singular quantitative measure of financial integration, most observers agree that it is closely associated with financial openness and capital mobility. Figure 1 depicts gross and net financial flows in and out of emerging market economies in Asia, Europe, and Latin America. Mirroring the global trend in financial and capital account deregulation that occurred during the 1980s and 1990s, capital flows into emerging market economies surged and then peaked just prior to the global financial crisis of Asia has become a major destination for such capital flows, accounting for nearly 60% of total 1 In this paper, the phrase emerging market economies refers to the emerging economies of Asia, Europe, and Latin America. Similarly, emerging Asia" refers the People s Republic of China (PRC); Hong Kong, China; India; Indonesia; the Republic of Korea; Malaysia, the Philippines; Singapore; Taipei,China; Thailand; and Viet Nam. Emerging Europe thus refers to Belarus, Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Moldova, Poland, Romania, the Russian Federation, the Slovak Republic, and the Ukraine, and emerging Latin America refers to Argentina, Brazil, Chile, Columbia, the Dominican Republic, Ecuador, Guatemala, Mexico, Peru, and Venezuela.

11 4 І ADB Economics Working Paper Series No. 351 financial flows into emerging market economies during the period prior to the global financial crisis. Figure 1: Financial Inflows and Outflows In and Out of Emerging Asia, Europe, and Latin America, Inflows 1000 US$ billion ,000 1,500 Outflows Emerging Europe Emerging Latin America Emerging Asia Net Financial Flows Note: Emerging Asia includes the People's Republic of China; Hong Kong, China; India; Indonesia; the Republic of Korea; Malaysia; the Philippines; Singapore; Taipe,China; Thailand; and Viet Nam. Emerging Europe includes Belarus, Bulgaria, Czech Republic,Estonia, Hungary, Latvia, Lithuania, Moldova, Poland, Romania, Russian Federation, Slovak Republic, and Ukraine. Emerging Latin America includes Argentina, Brazil, Chile, Colombia, Dominican Republic, Ecuador, Guatemala, Mexico, Peru, and Venezuela. Emerging Markets includes countries in Emerging Asia, Europe and Latin America. Data on inflows are liabilities, while outflows are assets. Foreign direct inflows from refer to direct investments in the reporting country. Other investments include financial derivatives. Data from 2005 onwards follow Balance of Payments Manual 6 (BPM6). Source: Author's calculations using data from International Financial Statistics, International Monetary Fund; and national sources. There are several ways to quantitatively assess financial openness, and thence financial integration. The first of these would be to count the number of legal restrictions on cross-border capital flows applicable to the economy concerned. This would constitute a de jure measure of financial integration, since it is based on legal restrictions currently in force. A wide variety of such restrictions are currently in use, including price- and quantity-based controls on the international movement of capital, as well as restrictions on equity holdings by nonresidents. Indeed, the International Monetary Fund reports more than 60 different types of such controls in its Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER). The inherent weakness of this approach is that counting capital controls gives us no information whatsoever regarding the degree to which such restrictions are effective in limiting the extent of international capital flows. A second approach would be to use an actual measure of financial openness, such as assessing actual cross-border capital flows, since the latter directly reflect the degree of financial openness of the economy concerned. In many ways, such an approach constitutes an improvement over the legal approach, in that, what matters is not how open capital markets

12 Asian Capital Market Integration І 5 appear on paper, but how integrated they are in practice. However, such a measure raises the question as to whether gross or net financial flows constitute the more suitable measure, particularly since both tend to be volatile and often subject to procedural issues in their measurement. Because of such shortcomings, the stock of foreign assets and liabilities (IMF 2009) is often used as an additional measure of actual financial openness and then financial integration. Figures 2a c depict de jure and de facto financial integration indexes for emerging Asia and its various subgroupings. The measure of legal financial integration used is the index initially developed by Chinn and Ito (2006) to measure capital account openness, which assesses the existence of legal restrictions on cross-border financial transactions as reported in the AREAER, values of greater magnitude indicating a greater degree of financial openness. Conversely, the summed gross stock of foreign assets and liabilities as a percentage of gross domestic product is the actual measure of financial integration depicted in Figure 2a c. The values depicted in Figure 2 were calculated from an update of the dataset constructed by Lane and Milesi-Ferretti (2007), and were extended with data from the International Monetary Fund s (IMF) International Investment Position Database. As depicted in Figures 2a c, emerging Asia's legal financial integration index rose sharply during the late 1970s and early 1980s, reflecting the wave of financial deregulation that occurred across the region during this period. Although there were some adjustments to legal restrictions relating to financial integration in the wake of this big bang style of financial deregulation that occurred during the 1970s and 1980s, the legal financial integration index remained high up until the late-1990s, the period just prior to the Asian financial crisis. The fall in the legal financial integration index around the time of the Asian financial crisis reflects increased foreign exchange interventions by Asian economies in the run-up to the crisis of 1997, as well as the introduction of capital control measures following the outbreak of the crisis. Notably, the trends in the legal financial integration indexes for the newly industrializing economies (NIEs) and the ASEAN-4 countries diverge quite widely in the wake of the crisis of In particular, while the NIEs quickly dismantled the temporary capital control measures put into place in the wake of the crisis and continued on their medium-term path of financial deregulation, the ASEAN-4 countries sustained their post-asian financial crisis controls overall, and even introduced additional controls during the global financial crisis of Interestingly, the actual financial integration indexes for all three sets of countries emerging Asia overall, the ASEAN-4, and the NIEs show a steady uptrend despite the declines that occurred in the legal financial integration index in the ASEAN-4. This highlights the substantial gap between de facto and de jure measures of financial openness and integration. In short, the fact that numerous developing countries impose significant legal controls on the international movement of capital does not mean that such controls are particularly effective in curbing international capital flows. Actual financial integration as measured by capital flows and cross-border financial asset holdings may significantly exceed the level implied by corresponding legal controls. This seems particularly relevant to the ASEAN-4 countries, where de facto financial integration measure increased rather sharply during the period following the 1997 crisis, just when the de jure measure of financial integration fell significantly. 2 As used in this chapter, the phrase "newly industrializing economies" (NIEs) refers to Hong Kong, China; the Republic of Korea; Singapore, and Taipei,China. Similarly, the phrase "ASEAN-4" refers to Indonesia, Malaysia, the Philippines, and Thailand, these countries being four of the ten members of the Association of Southeast Asian Nations (ASEAN). At this writing, ASEAN's full membership includes Brunei Darussalam, Cambodia, Indonesia, the Lao People's Democratic Republic, Malaysia, Myanmar, the Philippines, Singapore, Thailand, and Viet Nam.

13 6 І ADB Economics Working Paper Series No. 351 Figure 2: De Jure and De Facto Financial Integration in Emerging Asia, % of GDP Figure 2a: Emerging Asia Index % of GDP Figure 2b: ASEAN Index % of GDP Figure 2c: Newly Industrializing Economies Index De facto (LHS) De jure (RHS) Note: The de facto (LHS = left hand scale) measure that appears in Figures 2a-c refers to the sum of foreign assets and liabilities expressed as a percentage of GDP. This measure is taken from the Lane and Milesi-Ferreti External Wealth of Nations dataset (available at and was extended using data from the International Monetary Fund's (IMF's) International Investment Position (IIP). The de jure (RHS = right hand scale) measure refers to the estimated Chinn-Ito Index, which is based on the IMF's Annual Report on Exchange Arrangements and Exchange Restrictions (available at hhtp://web.pdx.edu/~ito/chinn-ito_website.htm). The Emerging Asia grouping includes the People's Republic of China, India, Indonesia, Malaysia, the Philippines, and Thailand. The Newly Industrializing Economies (NIEs) grouping includes Hong Kong, China; the Republic of Korea; Singapore, and Taipei,China. Source: External Wealth of Nations Dataset and the Chinn Ito Index.

14 Asian Capital Market Integration І 7 Park and Lee (2011) report that capital flows in and out of emerging Asia have consistently increased, and that over the past decade these have been particularly driven by portfolio investment flows. Their study likewise analyzed international portfolio asset holdings using the IMF s Coordinated Portfolio Investment Survey (CPIS), which disaggregates each economy's stock of portfolio investment assets by country of residence of the issuer of the asset concerned 3 Park and Lee find a sharp increase in emerging Asia s international portfolio asset holdings over the study period, implying a significant increase in the degree of financial openness and integration achieved by the region overall. However, when disaggregated by subregional grouping as well as asset classification, their assessment of the region s foreign asset holdings suggests that the degree of financial integration achieved by the region's various subgroupings of economies and market segments is far from uniform. Such findings suggest that there remains considerable room for improvement in regional financial integration in two ways. First, although emerging Asia s foreign portfolio assets are increasingly held within the region, advanced economies still account for a major share of its foreign portfolio assets. Second, there is a considerable difference in the degree of financial market development and integration as regards the equity and bond markets. In particular, Park and Lee find that the region s equity markets are relatively more open and integrated as compared to its domestic currency bond markets. Emerging Asia s foreign asset holdings are also skewed toward equities as opposed to debt securities. Further, the region s domestic currency bond markets remain largely segmented, their limited integration probably reflecting a correspondingly limited level of development of this segment of the financial market. III. CAPITAL MARKET INTEGRATION IN EMERGING ASIA De facto measures of financial integration can be grouped into two broad types: price- and quantity-based measures. Quantity-based measures reflect actual financial flows and crossborder asset holdings, as previously discussed. Conversely, the theoretical foundation for pricebased measures is the law of one price, which holds that given full financial integration, markets price assets with similar risk characteristics in a similar manner. In other words, once risk has been fully accounted for, the greater the degree of financial integration, the more closely correlated the movements of prices of assets of similar risk profiles. A. Integration of Stock Markets in Emerging Asia A substantial body of literature reports that stock price movements tend to be correlated with one another over time, international boundaries notwithstanding. Further, this correlation tends to be closer during market downturns and periods of financial turmoil than during other periods (King and Wadhwani 1990, Longin and Solnik 1995, Karolyi and Stulz 1996, and Forbes and Rigobon 2002). This gives rise to a natural concern that financial crises raise the degree of correlation between asset price movements across international borders, thus increasing the possibility of contagion of financial turmoil from one domestic market to another. This increased correlation between asset price movements across international borders occurred in emerging Asian markets in the aftermath of the global crisis of 2008, as regional and global cross-border asset price correlations tended to be closely associated during that period (Park and Lee, 2011). Hinojales and Park (2010) also find that the degree of integration of emerging Asia stock 3 The first Coordinated Portfolio Investment Survey (CPIS) dataset which was published in 1998 reflected data for However, annual CPIS data only became available in 2001.

15 8 І ADB Economics Working Paper Series No. 351 markets has increased over time, thereby suggesting increased risk of contagion from one domestic market to another. This section of the paper assesses the degree to which movements in stock-market returns in emerging Asian economies are correlated with those at the regional and global level. Individual economy, regional, and global stock price indexes are used for this purpose. The stock price indexes of individual economies used in performing the analysis are those computed by Datastream International, these indexes including firms of large, medium, and small market capitalization. The dataset includes 10 emerging Asia economies and 19 advanced countries The 10 emerging Asia economies include the People's Republic of China (PRC); India; Indonesia; the Republic of Korea; Malaysia; the Philippines; Singapore; Taipei,China; and Thailand. The advanced economies include Austria, Australia, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Portugal, Spain, Sweden, Switzerland, the United Kingdom (UK), and the United States (US). Financial returns to stocks are expressed in local currency units, with a weekly frequency beginning 13 January 1993 and ending 28 December These weekly data refer to the mid-week (Wednesday) closing price for each stock concerned. Weekly stock market returns are computed as the difference between the current and previous week s closing price expressed in percent. Ultimately, the purpose of the study is to measure the degree to which financial returns in emerging and advanced countries converge over time, as such convergence indicates financial integration. To do this, the study employs two notions of convergence. The first is σ- convergence, which uses the filtered cross-sectional dispersion of financial returns in individual economies following the methodology used by Adam et al. (2002) and European Central Bank (2004). This measure indicates an increase in the degree of convergence of financial returns across economies (and thence financial integration) by a decrease in the cross-sectional standard deviation σ of financial returns across study economies. Full integration is thus achieved when the cross-sectional distribution ends in a single point, and the standard deviation approaches zero. Figure 3 graphically depicts changes over time in σ-convergence in stock market returns across the study, economies included in the emerging Asia grouping and its subgroupings. As seen in the figure, the cross-market dispersions of weekly stock market returns have declined over time for both the emerging Asia grouping and its subgroupings, thus suggesting increasing integration of the stock markets in the emerging Asia grouping and its subgroupings. The same holds true for the cross-market dispersion of weekly stock market returns in emerging Asia and its subgroupings as compared to the average of stock market returns globally. It is noteworthy that this dispersion declined to a greater degree for the newly industrializing economies (NIEs) than for the ASEAN-4 countries. Interestingly, the cross-market dispersion of weekly stockmarket returns in emerging Asia increased during the crisis periods of and , reflecting the impact of these crises on individual market returns, as well as on asset price volatility. 4 In order to net out the impact of changes in the exchange rate, the stock and bond market returns used in performing the analysis are expressed in domestic-currency terms. This was deemed an important feature of the analysis, since the magnitude of variation in weekly returns is generally small as compared to the magnitude of changes in the exchange rate, especially in the case of bond markets. While some financial market integration studies use financial returns denominated in US dollars, the impact of exchange rate variations on the results achieved remains a source of debate.

16 Asian Capital Market Integration І 9 Figure 3: σ-convergence of Weekly Stock-Market Returns for Selected Groupings of Economies (September 1991 November 2011) Sep 91 Mar 94 Sep 96 Apr 99 Oct 01 Apr 04 Nov 06 May 09 Nov 11 EmergingAsia ASEAN-4 NIEs World Note: The values depicted on the x-axis refer to the standard deviation of the filtered country returns obtained by employing the Hodrick-Prescott method, with the parameter (lambda) set to a value of 270,400. The Emerging Asia grouping refers to the People's Republic of China; Hong Kong, China; India; Indonesia; the Republic of Korea; Malaysia; the Philippines; Singapore, Taipei,China; and Thailand. The Newly Industrializing Economies (NIEs) grouping refers to Hong Kong, China; the Republic of Korea; Singapore; and Taipei,China. The ASEAN-4 grouping includes Indonesia, Malaysia, the Philippines, and Thailand. The World grouping includes 34 advanced and emerging economies including those in the Emerging Asia grouping identified above. The stock price indexes for all individual economies included in the analysis are expressed in units of domestic currency. All stock-price-index data were downloaded from Datastream. Source: Author's calculations using data accessed from Datastream accessed July The second measure of convergence employed by the study is β-convergence, which was first used in growth theory, but was then later applied to financial integration by Adam et al. (2002) and Rizavi et al. (2011). Using the specification employed by Rizavi, Naqvi, and Rizvi (2011), β is estimated by the equation: L ERit, it, ERit, 1 i 1 1 ERit, 1 it, (1) where represents the difference between the return in economy i and the computed regional return (rr), and Δ, is the difference between the returns in two successive periods, and ER R R (2) it, it, rrt, where R i,t refers to weekly returns computed as the natural log difference between the value of the current and previous stock (or bond) price index. The regional returns are calculated using principal component analysis. A negative value for β suggests that the values of the returns converge, the magnitude of β indicating the rate at which convergence (i.e., financial integration) takes place. As explained by Rizavi, Naqvi, and Rizvi (2011), the β-convergence coefficient in [equation (1)] can take on a value ranging from 0 to 2. A negative sign for the coefficient indicates financial integration, with the highest possible speed of convergence (financial integration) being indicated by a value of 1. The extreme values 0 and 2 indicate no financial integration whatsoever. Conversely, a value for β between 0 and 1 indicates gradual unidirectional

17 10 І ADB Economics Working Paper Series No. 351 convergence between the returns in the two markets concerned, while a value for β between 1 and 2 implies that the convergence is oscillating or fluctuating. Given that financial returns vary over time, it is possible for the estimated values for β- convergence to change over time. To address the problem of β-convergence changing over time, the study uses a rolling estimation procedure following Fratzscher (2001). This procedure estimates β over a fixed 1-year (52-week) period beginning January 1993 and ending in January This 1-year window is then moved forward 1 week at a time until the final observation in the dataset is included in the calculation. Figure 4 depicts the β-convergence coefficients for weekly equity market returns for individual economies as compared to the returns for the emerging Asia grouping as a whole, the latter being calculated as the unweighted average of βs for all economies in the emerging Asia grouping. The negative values assumed by the βs indicate convergence in individual economy equity market returns with the equity market returns at the regional level, those these βs oscillate over time. This oscillating pattern of convergence is notable during the crisis periods of and , as it indicates a significant slowdown in the convergence process, though this slowdown quickly reversed during and , the years directly following the two crises. The NIEs show the greatest degree of convergence, followed by the ASEAN-4 grouping. The convergence between the weekly equity-market returns for the PRC and the emerging Asia grouping as a whole takes place in a gradual manner. Figure 4: β-convergence of Individual-Economy Weekly Equity Market Returns with the Weekly Equity Market Returns of the Emerging Asia Grouping as a Whole Jan 94 Jul 96 Feb 99 Aug 01 Mar 04 Sep 06 Mar 09 Oct 11 PRC IND KOR PHI TAI HKG INO MAL SNG THA Emerging Asia PRC= People s Republic of China; HKG = Hong Kong, China; IND = India; INO = Indonesia; KOR = Republic of Korea; MAL = Malaysia; PHI = Philippines; SNG = Singapore; TAI = Taipei,China; THA = Thailand. Note: The values for the emerging Asia correspond to the unweighted average of the β-convergence coefficients for each of the individual economies included in the emerging Asia grouping. The regional returns for each member country are computed using the first principal component analysis. Source: Author s calculations.

18 Asian Capital Market Integration І 11 B. Integration of Bond Markets The present study also assesses the degree of convergence between individual economy, local currency bond yield indexes and the corresponding index for the emerging Asia groupng as a whole, as well as the former with the composite global bond yield index. The data used pertain to the zero coupon yield curve for each economy under study, since this nets out differences in coupon rates, maturities, and individual bond idiosyncrasies across economies, thus allowing a straightforward comparison of bond yields across individual economies and groupings (Gürkaynak, Sack, and Wright, 2007). Data for the bond yields pertaining to individual economies are the relevant benchmark yields sourced from Bloomberg LP. Figure 5 depicts the σ-convergence of composite 10-year government bond yields for (i) the NIEs, (ii) the ASEAN-4 grouping, and (iii) the emerging Asia grouping, all of these being compared with US Treasury bond yields of the same tenor. Since the year 2000, the dispersion of bond yields for the ASEAN-4 has declined significantly as compared to the other groupings, while the NIEs dispersion of bond yields remained limited as compared to the emerging Asia grouping as a whole. Interestingly, the dispersion of bond yields for the emerging Asia grouping as a whole declined to a greater extent than did the dispersion of the bond yields for the global grouping since late This implies growing convergence among bond yields for the emerging Asia grouping. In contrast, the dispersion of bond yields for the overall global grouping has increased since 2010, as some countries included in this grouping have encountered sovereign debt problems. Figure 5: σ-convergence of 10-Year Government Bond Yields % Aug 00 Dec 01 May 03 Sep 04 Feb 06 Jul 07 Nov 08 Apr 10 Aug 11 Emerging Asia Notes: 1. Values refer to standard deviation of country bond yield spreads over 10-Year US government bond yields. Data are filtered using Hodrick-Prescot tmethod. 2. Emerging Asia includes the People's Republic of China; Hong Kong, China; India; Indonesia; the Republic of Korea; Malaysia; the Philippines; Singapore, Taipei,China; and Thailand. 3. ASEAN-4 = Association of Southeast Asian Nations includes Indonesia, Malaysia, the Philippines, and Thailand. 4. NIEs = newly industrialized economies include Hong Kong, China; the Republic of Korea, Singapore; and Taipei,China. 5. World includes 26 advanced and emerging economies including those from Emerging Asia. Source: Author's calculations using data accessed from Bloomberg LP accessed July ASEAN-4 NIEs World

19 12 І ADB Economics Working Paper Series No. 351 Figure 6 presents the σ-convergence of bond yields for the emerging Asia grouping as compared with US Treasuries with the same tenor for varying maturities. It shows that the dispersion of government bond yields for various maturities has declined since late 2005, although this trend was interrupted during the crisis years of Notably, the dispersion of yields across the various maturities was of roughly equal magnitude in 2005, implying that the degree of dispersion became similar over time across tenor. Figure 6: σ-convergence of Emerging Asia Government Bond Yields for 2-Year, 5-Year, and 10-Year Maturities % Aug 00 Dec 01 May 03 Sep 04 Feb 06 Jul 07 Nov 08 Apr 10 Aug Year 5 Year 2 Year Note: Values refer to the standard deviation of country bond yield spreads over 10-Year, 5-Year, and 2-Year US government bond yields. Values are filtered using Hodrick-Prescot t method with parameter λ set to 14,400. Emerging Asia includes the People's Republic of China; Hong Kong, China; India; Indonesia; the Republic of Korea; Malaysia; the Philippines; Singapore, Taipei,China; and Thailand. Source: Author's calculations using data accessed from Bloomberg LP accessed July Figure 7 shows the β-convergence of weekly coefficients of the unweighted regional average and individual-economy five-year government bond yields as compared with computed regional yields (derived using principal component analysis). Several points regarding Figure 7 are noteworthy. First, the β-convergence coefficients approach zero during the sample period. Although they are generally negative in sign, which suggests some convergence in bond yields, their magnitudes are much smaller as compared to the corresponding values for equities. In fact, for some economies, the coefficients turn positive at some points, indicating divergence. Second, the convergence of individual economy bond yields with regional yields has been consistently gradual over time, as the values for the βs range between 0 and 1. Third, the degree of convergence among regional bond yields increased significantly during the global financial crisis, though this sharp increase in convergence appears to have been transitory.

20 Asian Capital Market Integration І 13 Figure 7: β-convergence of Individual-Economy 5-Year Government Bond Yields with Regional Bond Yields Jun 06 Nov 07 Mar 09 Aug 10 Dec 11 PRC IND KOR PHI TAI HKG INO MAL SNG THA Emerging Asia PRC= People s Republic of China; HKG = Hong Kong, China; IND = India; INO = Indonesia; KOR = Republic of Korea; MAL = Malaysia; PHI = Philippines; SNG = Singapore; TAI = Taipei,China; THA = Thailand. Note: The values for the emerging Asia correspond to the unweighted average of the β-convergence coefficients for each of the individual economies included in the emerging Asia grouping. The regional yields for each member country are computed using the first principal component analysis. Source: Author's calculations. C. Degree of Financial Integration and Extent of Spillover of the Negative Impacts of Regional and Global Financial Crises into Emerging Asia Economies Financial integration with countries both within and outside the region may increase the degree to which emerging Asia economies are negatively impacted by external financial shocks. However, direct evidence of this remains incomplete at this writing. What can be said with relative certainty is that given that emerging Asia equity and bond markets are fully integrated with their counterpart markets globally, and given absence of financial disturbances specific to emerging Asia economies, then the only information to which the stock prices and bond yields in emerging Asia markets will respond is that same information that impacts all markets globally. By extension, analyzing the reaction of these markets to regional and global shocks allows us to assess the extent to which emerging Asia equity and bond markets are financially integrated with markets within and beyond the region. 1. The Data For the individual economy stock market returns included in the study, the same data as those referred to in Section III. A were used. These include the stock price indexes of individual economies computed by Datastream International, which encompasses firms of large, medium,

21 14 І ADB Economics Working Paper Series No. 351 and small market capitalization. The dataset includes 10 emerging Asia economies and 19 advanced countries. The 10 emerging Asia economies include the PRC; India; Indonesia; the Republic of Korea; Malaysia; the Philippines; Singapore; Taipei,China; and Thailand. The advanced economies include Austria, Australia, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Portugal, Spain, Sweden, Switzerland, the UK, and the US. Financial returns to stocks are expressed in local-currency units with a weekly frequency beginning 13 January 1993 and ending 28 December These weekly data refer to the mid-week (Wednesday) closing price for each stock concerned. Weekly stock returns are computed as the difference between the current and previous week s closing price expressed in percent. For the indexes of returns to bonds, the HSBC Asia Local Bond Index (ALBI) was used for economies in the emerging Asia grouping, and the Citigroup World Government Bond Index for all others. Both datasets are denominated in domestic currency, with a weekly frequency beginning on 13 January 1993 and ending on 28 December 2011 for stocks, and from 10 January 2001 to 28 December 2011 for bonds. 6 While price-based measures such as those used by the present study are often useful, they present some practical problems. In particular, returns to financial assets may be subject to a multitude of risk and liquidity premiums that are difficult to quantify. Further, the pricing mechanism in emerging-market financial markets may not function in an efficient manner. This is particularly true in shallow emerging-economy financial markets where trading is thin. Further, given that such markets are fully integrated globally, domestic asset prices may be particularly impacted by global and regional financial distress. Moreover, in such cases, empirical results are often swayed by the benchmark equity and bond indexes selected. Thus, if selection of the latter is inappropriate, the results may be misleading. Earlier studies have often included unexpected returns on regional and global indexes of particular asset classes. However, these regional or global indexes themselves may not be the best benchmark index for representing regional or global assets. An important question for the present study is whether or not there is a common feature among Asian asset prices that can capture co-movement of prices at the regional level. One way of addressing this issue is to use regional price indexes as benchmark indexes. However, these are often simply weighted averages of individual economy stock and bond market indexes. As a result, they do not necessarily capture co-movement of Asian asset prices. The present study uses principal component analysis to create benchmark indexes of returns to stocks and bonds at the regional level. Principal component analysis yields principal components, capturing as much information common to the variables concerned as possible. The first principal component has the largest possible variance, and thus accounts for as much 5 6 In order to net out the impact of changes in the exchange rate, the stock and bond market returns used in performing the analysis are expressed in domestic currency terms. This was deemed an important feature of the analysis, since the magnitude of variation in weekly returns is generally small as compared to the magnitude of changes in the exchange rate, especially in the case of bond markets. While some financial market integration studies use financial returns expressed in US dollars, the impact of exchange rate variations on the results achieved remains a source of debate. Both datasets include 10 emerging Asian economies (the PRC; India; Indonesia; the Republic of Korea; Malaysia; the Philippines; Singapore; Taipei,China; and Thailand), as well as advanced countries including Austria, Australia, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. The equity dataset also includes Argentina, Chile, Mexico, South Africa, and Turkey. The dataset for analyzing the bond market also includes Greece and Norway.